1. Field of the Invention
This invention relates to a network of computer systems, including but not limited to the Internet environment, and more specifically for authenticating a printer in the network environment.
2. Description of the Related Art
A network environment can comprise an endless number of configurations, including but not limited to computer systems communicatively connected to the Internet, to a wide area network, to a local area network, using TCP/IP connections, using token ring connections, etc. Likewise, the computer systems themselves may vary from network terminals with minimal storage and CPU processing functionality to personal computers including laptop computers to workstations to servers to mainframes. The relationship among the computers can vary, e.g., as being independent from each other, or having distributed relationships, or having client/server relationships, etc. Some or all of the files may be stored in a dedicated file storage system, e.g., a file server, database management system, etc., or within the storage of each system. Likewise, printers may be attached to any or all of the systems and/or there may be print servers to which the computer systems can be communicatively linked.
There are many different types of security issues that arise in a network environment. Some files must be encrypted at the sending end and decrypted at the receiving end to ensure that the file contents are not intercepted by an unauthorized entity during the transmission. This security feature, along with other security features that are known, can guarantee that a file has not been tampered with or can ensure the identity of the sender or receiver. Some of these security features are further discussed below.
Cryptography
Conventional cryptography, or in other words traditional symmetric cryptography, is used to maintain the privacy of the information contents. Conventional cryptography requires that the sender and receiver of an encrypted message share the same secret key. The same key is used to both scramble (encrypt) and unscramble (decrypt) information. In 1977, the National Bureau of Standards approved a block cipher algorithm referred to as the Data Encryption Standard (DES). Binary-coded data is protected by using the DES algorithm in conjunction with a key. An authorized user must have the key that was used to encipher the data in order to decipher it. Unauthorized recipients of the ciphered information content who may know the DES algorithm but who do not know the key cannot decipher the information content.
The major problem with this method is guaranteeing that both sender and receiver have the key but no one else does. Sharing the key requires that one party send it to the other. However, since most communication networks cannot be trusted, the key itself must be encrypted. If it is sent in the clear, there is a danger that someone eavesdropping on the line could get the key and then be able to decode messages sent between the two parties. Others have sent the key via registered mail, which slows the communication process down, and begs the question of why not just send the message registered mail if time is not of the essence.
As described above, to protect the information content from unauthorized recipients, the key has to be kept secure from unauthorized users. Thus, the security of the contents depends upon the security of the key. As such, the key has to be distributed to authorized users in a secure manner.
Public Key Cryptography
Public key cryptography was first introduced by Whitfield Diffie and Martin Hellman of Standford University in 1976. It not only can be used to ensure the privacy of transmitted messages, but it can also be used in other applications, including digital signatures.
For ensuring the privacy of transmitted messages, public key cryptography does solve many of the problems, discussed above, of securely distributing the key used in conventional cryptography.
Public key cryptography is based on two keys, a private key and a public key, that work together. A person's public key is openly made available to others, while their private key is kept secret. One key is used for ciphering and the other key is used to decipher information content. For each encryption key there is a corresponding, but separate and distinct, decryption key. Messages encrypted with a person's public key can only be decrypted with that person's private key. Even if one key is known, it is not feasible to compute the other key.
In a public key system, it is possible to communicate privately without transmitting any secret key. For example, the encryption key for each user is made public by being distributed or published. Anyone desiring to communicate in private with a recipient merely encrypts the message under the recipient's public key. Only the recipient, who retains the secret decrypting key, is able to decipher the transmitted message.
A combination of conventional cryptography and public key cryptography allows a secret key to be sent securely to an intended recipient. The sender encrypts a message with the secret key using the recipient's public key. The recipient then uses the recipient's private key to decrypt the message and to get the secret key for other transmissions. Since public key encryption is slower than secret key encryption, this approach allows subsequent transmissions to use the faster conventional secret key cryptography approach.
Digital Signatures
In these cryptographic systems, there is sometimes still a need to verify that the sender of a received message is actually the person named in the message. Digital signatures, which are based on public key cryptography, are used as a means to authenticate the sender of a message. A digital signature allows a digital message to be signed so that any receiver of a digitally-signed electronic message can authenticate the sender of the message and verify the integrity of the signed message. That is, the recipient is assured that the message is received as sent, and that it is not a forgery.
To ensure that the original true sender sent the message, a process just the opposite of the one used to ensure a private communication using public key cryptography described above is used. For example, a user who has made public a public key can digitally sign a message by encrypting the message, or a hash of it, with the user's private key before transmitting the message. Recipients of the message can verify the message or signature by decrypting it with the sender's public encryption key. This process is just the opposite of conventional cryptography in that the message is first encrypted by the sender using the sender's private key and decrypted by the recipient using the sender's public key. Anyone who has the sender's public encryption key can read the message or signature. Any such recipient is assured of the authentication of the creator of the message since only the sender having the secret private key could have created the message or signature. The recipient is also assured that the message has not been altered since it was first created and the digital signature was attached to it. Any recipient can authenticate the digital signature and verify the integrity of the message by using only the signer's public key.
In the above example, the digital signature was the encryption, using the sender's private key, of the message itself. In the Digital Signature Standard (ANSI X9.30 Part I) a person's digital signature is a fixed-length string of bits that are attached to an electronic message of any length. To create a fixed-length digital signature, a hashing function is used that converts a message of any length to the same fixed-length hash, or digest, of the message. The Secure Hash Algorithm (SHA) is a known hash function that is part of the Digital Signature Standard. This hash of a message is like a “fingerprint” in that it is practically impossible for two distinct messages to result in identical hashes. After creating a hash of the message, the sender's private key is applied to the hash to create the digital signature for the message. The digital signature is a function of both the message being signed and the signer's private key. As long as the private key is kept secret, the digital signature cannot be created by anyone else.
Upon receipt of the digitally-signed message, the recipient uses the sender's public key to convert the digital signature to the hash that the sender computed. Next, the recipient applies the same hash function to the plain text message received and gets the hash of the received message. If the hash of the received message is identical to the hash obtained by using the sender's public key to convert the digital signature, then the recipient has authenticated the sender's digital signature and verified the integrity of the signed message.
Certificates
The identity of the signer can only be guaranteed to the extent that the receiver is assured that the public key actually belonged to the purported sender. One known technique for addressing this problem is to rely on some trusted authority, e.g., a government agency, to ensure that each public key is associated with the person claiming to be the owner. The trusted authority would create a digital message, known as a certificate, which contains the claimant's public key and the name of the claimant. A representative of the authority would sign the digital message with the authority's own digital signature. The authority's digital signature would be created by using the authority's private key, and it would be deciphered by recipients using a public key of the authority which has been widely disseminated and made available such as through telephone books, newspapers, and/or on an Internet web page. This certificate is sent along with the sender's message and the sender's digital signature. The recipient uses the authority's public key to decipher the certificate and find the sender's authentic and certified public key. The recipient then uses the sender's certified public key to verify the sender's signed message. Thus, the certificate can be easily authenticated and the message integrity verified.
Access Control Using Certificates
Typically, access to resources of a computer system (“server”) from another system or user (“user”) has been controlled through passwords. This requires the server to maintain a database of all authorized users and each user's password. However, if a user shares the password with another unauthorized user, the integrity of the password access control system is diminished.
In a certificate-based access control system, the server only needs to authenticate certificates issued by a certification authority. The server does not need to maintain a database about users or each user's corresponding password. To gain access to resources of the server, the user submits the user's certificate. From the certificate, which contains unforgeable data, the server can obtain the user's authenticated public number, personal data, and access privileges. The server can then transmit to the user a random message that the user must digitally sign with the user's private number and return it to the server. The server can then authenticate the digital signature using the public number in the certificate and check that the signed message is the same it sent to the user. With this digitally-signed response, the server can determine if the user has the correct private number corresponding to the authenticated public number in the certificate.
Secured Transmissions Between a Sender and Receiver
The above-described secure transmission techniques are best applied in situations where the messages and/or files are transmitted directly between the sender and the intended user.
In any network environment, situations may arise where a user (an individual interacting with a system via a terminal or an application running on a system) desires to print a document that is located remote from the user.
The document may be protected from being accessed by anyone other than those users that have access privileges.
Typically, a user will request the document from the remote system, the remote system will verify that the user has the correct access privileges, and if so, then the remote system will send a copy of the document to the user. The user will then send the file to a printer for printing. However, such a user having access privilege may desire to print the document on a remote printer or print server but does not desire to first retrieve and store the document at the user's own local computer system (referred to for convenience as the client system). For various reasons, a user may not wish to have the document resident on the user's own machine. Some of these reasons may involve, for example, any one or more of the following: the client system may not be in a secure environment; there may be network traffic considerations; or the client system may not have the storage space for receiving the file, etc. In addition, the file server may not want a copy of the file to be stored on the client system. The owner of the file (e.g., document) may wish to control the number of copies being distributed, e.g., to protect copyright in the document and/or payment of a fee on a per-copy basis. If a copy were resident on the client machine, illegal copies could be further made from that copy, or illegal changes could be made to the document. Instead, it may be more desirable if the printer could get the document directly from wherever it may be stored and print the document.
However, in order to do this, the printer would need to have the same access privileges as the user had if the document was access protected.
There is a need to allow a print server to get a print file from a third party identified in an original request so that the document can be printed without first obtaining the file by a client system originally requesting the print file. However, when the print server gets the file, the third party must be guaranteed that the request is valid (i.e., the print server has been authorized to get the file, and the original client can legally print the document). Such a scenario is not known to be possible under existing protocols.
Printer Security
There are security issues involved with printing, and even more so when the printing involves network communication links that are not secure, e.g., the Internet. One aspect of more secure printing involves granting administrative access to a device, such as a printing device, on the network. Several vendors provide tools to aid in the installation and configuration of network attached printers. These tools allow a user to configure the printer, set defaults, and even download flash memory. These are powerful and necessary tools for an administrator dealing with the management of large networks of printers. However, in the wrong hands, these tools can be used to completely disable a printer, or potentially damage the entire network. Therefore, printers, themselves, need to be protected from malicious attacks. Simple ID and password protection may not be sufficient, especially if a printer is accessible from the Internet or from outside a firewall.
Since it is now common for critical data to be printed via a network that is not totally within a user's own controlled boundary, the user needs to know that the printer, or printing system, or other output device is the one that the user believes it to be. For example, when printing the payroll, one would like to be assured that the payroll checks are really going to a specific printer, that the data has not been tampered with, and that an intruder has not captured the data. It is desirable for a user to be able to authenticate a printer, i.e., validate that the printer is really the printer that it claims to be.
Currently, there is a significant security exposure in printers attached to networks in that there is no way to authenticate that a printer really exists at a specified network address. That is, software could be used to fool the system or a potential user into thinking that a printer was at a network address, but print data sent to that address could be misappropriated by hostile software and used for other purposes.
Although digital certificates are used in computing systems to authenticate parties who will trade information, currently secure printers, i.e., printers that are capable of authenticating themselves to a user or a foreign system, are nonexistent.
It is therefore an object of this invention to ensure that a printer, printing system, or other network device is the actual device the communicating sender assumes it to be.
It is therefore a further object of this invention to provide a secure means to configure a digital certificate in a network printer or other network device.
In order to enable a network device to operate in a secure way on a network, the system, method, and program of this invention provides a means for a network device to identify itself. The network device identifies itself in a way analogous to the way individuals identify themselves using a passport or driver's license that has been issued by a trusted authority. A trusted authority gives a certificate to the network device that guarantees that the device is who the device says it is.
Although the preferred embodiment herein involves a secure configuration of a printers digital certificate, the invention is applicable to other network devices including, but not limited to, fax machines, modems, personal digital assistants, cellular telephones, or other communication devices.
Using a printer as a specific example, when a printer is manufactured, a unique data encryption key is built into the printer. The unique key and the corresponding model and serial number (or other unique identifier) of the printer are maintained in a secure, network (e.g., web) accessible database server. For example, such a secure database server may be maintained by each printer manufacturer. For example, IBM would have a web site storing the keys for all new printers that IBM built.
When an administrator configures a printer, the administrator may also choose to configure a digital certificate. The printer provides a configure digital certificate option as part of the configuration process. When this option is selected, the following events occur:
First, the printer establishes a session (e.g., an HTTP session) with the database where the encryption key built into the printer is stored.
Second, the printer sends a two-part message to the server. The first part of the message contains the printer model and serial number, the printer's network address (e.g., IP address), and a request for a digital certificate. This first part of the message is encrypted with the encryption key built into the printer. The second part of the message contains the same data (information) as the first part, but the second part of the message is sent in the clear.
Third, the server decrypts the message, by using the printer's unique key in its database, and compares the decrypted message with the clear message. If both parts of the message match, then the server knows that the printer is authentic. No one else could have encrypted the message with the printer's unique key. The server also knows that the message came from the given network address.
Fourth, the server builds a digital certificate for the printer, registering itself as the certificate authority. The server then encrypts the content of the digital certificate with the certificate authority's private key, and sends it to the printer. Since the digital certificate is encrypted, it is safe to transmit. It should be noted that the content of the digital certificate conforms to the international standard, X.509, for digital certificates. As such, the digital certificate includes a new public/private key pair for the printer.
Fifth, the database at the certificate authority server stores the new public key just sent to the device. The public key is then advertised for this printer and is used in all subsequent secure transactions with this printer.
Sixth, the printer decrypts the digital certificate using the certificate authority's public key, and stores it in nonvolatile memory in the printing system.
Consequently, whenever a printer is requested to authenticate itself, it can do so by using the newly-configured digital certificate.